4.1.1. Cognitive effects

Concentration disruption increases progressively, with the greatest disruption occurring in the VF, followed by the VL, with the least disruption observed in the NV (Figure 3). This indicates that the presence of a voice and its meaningfulness disrupt the participants' concentration. The test results show that the degree of concentration disruption is greater for a meaningful voice compared to a meaningless voice. This suggests that VF significantly reduced concentration compared to NV, potentially contributing to the decrease in originality observed in Figure 3. While multiple hypotheses exist regarding the mechanism by which auditory stimuli cause task interference, Reference Mehta, Zhu and CheemaMehta et al. (2012) proposed that reduced information processing capacity due to auditory stimuli leads to a decline in creativity. In their experiment, it can be considered that some processing resources were automatically allocated to the meaningful voice, thereby reducing concentration on the creative task and causing an interference effect. Another discussion that refers to their study can be that the creativity promotion effect of increased abstract thinking was caused by diverted attention with moderate noise (Reference Mehta, Zhu and CheemaMehta et al., 2012; Reference OppenheimerOppenheimer, 2008). Some participants in this study answered that their attention was more diverted by the meaningful voice, though their originality scores had decreased. Consequently, there can be two possibilities: one is that abstract thinking was promoted by the diverted attention, but the effect was canceled out by the interference effect (similar to the high-noise condition observed in their experiment); the other is that, due to differences in tasks and auditory stimuli, the results differed.

4.1.2. Affective effects

NV and VL were rated as more conducive to a sense of calmness and relaxation compared to the VF. This finding suggests the possibility that such affective states influence originality. Numerous studies have reported that positive emotions enhance creativity. For example, in an experiment conducted by Reference Tang, Xia, Li, Wang, Ying and YangTang et al. (2023), participants were assigned a total of six trials of design tasks under three kinds of music conditions (positive music, negative music, and non-music). The results showed that participants produced more ideas in positive music condition and indicated that the positive music was useful for their performance in divergent thinking measures was significantly greater. Divergent thinking, the ability to generate multiple solutions or ideas for a given problem, is considered particularly important in creative tasks. They interpreted these results as an indication that positive emotions may broaden attentional focus, enabling more extensive information processing unconstrained by conventional patterns. It is conceivable that the calmness and relaxation induced by the non-voiced BGM in this experiment may have similarly contributed to higher originality scores.

4.2.1. Influence of temporal phases

Figure 4 indicates that no significant differences between any of the conditions were observed in fluency or originality during the first half of the experiment, whereas significant or near-significant differences in originality emerged during the second half. This suggests that the disparity between fluency and originality results likely arose in the latter half of the experiment. What distinguishes the first and second halves of the experiment? Oral interviews with participants revealed comments such as, “At first, I didn’t notice the lyrics, but as I struggled to come up with responses in the latter half, they started to catch my attention.” This indicates that, as the experiment progressed, the participants' attention began to shift away from the task and toward the auditory stimuli. This shift suggests that the influence of auditory conditions became more pronounced in the second half, with more cognitive resources being allocated to processing the sound. The results suggest that, as participants progressed into the latter half of the experiment, they generated fewer ideas overall but provided responses with more originality, often describing unconventional approaches. When asked how they came up with their answers during the interviews, participants' responses in the first half often included phrases like, “Because I’ve used it that way before,” reflecting general, familiar approaches. In contrast, responses in the second half were more likely to involve unique approaches derived from previously generated ideas. These observations imply that eliciting highly original responses in the latter half of the experiment is crucial to enhancing originality. As the experiment progresses, participants appear to draw more on unique perspectives and ideas, which likely contribute to the assessment of originality. However, producing such creative responses requires deeper cognitive processing. In the critical latter half, cognitive resources may have been diverted toward processing auditory stimuli, particularly under the VF and VL, resulting in lower originality scores under these conditions. For fluency, Figure 4 indicates that the highest scores were observed in the early stages of the experiment. This suggests that the quantity of responses generated during the initial phase is a critical factor for fluency. The lack of significant differences between conditions during this phase may be attributable to the smaller influence of auditory stimuli in the first half of the experiment.

4.2.2. Influence of cognitive modes

When participants were asked during interviews how they came up with their responses, many mentioned reasons such as, “I’ve used it that way before,” or “I remembered someone else using it that way.” These responses were likely based on retrieving prior knowledge or memories rather than creating something new, applied convergent thinking. These types of responses are naturally included in the fluency scores. In contrast, originality scores are based on the top four most-original responses, which are more likely to reflect novel ideas created by the participants. These responses align more closely with divergent thinking tasks. In their experiment, positive emotions did not enhance performance on convergent thinking tasks but significantly improved performance on divergent thinking tasks (Reference Tang, Xia, Li, Wang, Ying and YangTang et al., 2023). Similarly, in this study, the influence of positive emotions, such as the sense of calmness and relaxation, may have been more pronounced in the evaluation of originality, which primarily consists of responses that are closer to divergent thinking.

4.2.3. Influence of personal characteristics

A negative correlation was observed between fluency, originality, and music listening habits (Figure 5). This suggests that participants less familiar with music performed better on the tasks, which may seem counterintuitive at first glance. One might expect that those less accustomed to music would be more easily distracted by BGM, thus experiencing greater concentration disruption. However, this result can be explained by focusing on the performance enhancement effect caused by the positive emotions evoked by BGM. Specifically, the less familiar participants with music listening, the more significantly they are influenced by BGM and its vocal elements, the results can be interpreted as these participants more strongly benefiting from the performance improvement effect of positive emotions. Moreover, only in the daily music listening group did NV induce significantly higher performance than VF. This might seem to contradict the previous interpretation. Possible explanations could include the following: First, participants who listen to music daily might have adaptively processed BGM's influence, potentially showing a unique response to a meaningful voice. Alternatively, the daily music-listening group had approximately 1.5 times more participants than the non-daily group, which could simply have increased statistical detection power.

Both fluency and originality showed an overall positive correlation between Mind Wandering Questionnaire (MWQ) scores and creative task performance (Figure 6). This result is consistent with Reference Yamaoka and YukawaYamaoka et al.'s (2016) experiment, where the middle MWQ score group showed significantly more originality compared to the low-score group. They pointed out the relationship between mind wandering and cognitive disinhibition, arguing that “mind wandering is a state where attention deviates from the current task, and thoughts (information) automatically arise, which is similar to a state of cognitive disinhibition.” In this experiment, the strongest positive correlation between fluency and voiced conditions was observed with the meaningful voice, which may indicate a synergistic effect between a meaningful voice and the cognitive disinhibition aspect of mind wandering. As MWQ scores increase, cognitive filters weaken, making it more difficult to filter information. When a meaningful voice—a complex information source—is introduced, the abstraction of thought may be synergistically heightened, subsequently increasing fluency. Consequently, participants with greater mind-wandering tendencies might have experienced particularly enhanced fluency, creating a wider performance gap from those having lower such tendencies, thus resulting in a positive correlation.